This preregistration is intended as a supplement to a previous preregistration that is already stored as part of the current OSF project (https://osf.io/wc6sv/overview). In the original preregistration, we committed to several planned analyses. The present preregistration builds on that initial document and extends it in several ways. First, in the present analysis, we will focus exclusively on the affective-report condition, in which participants reported or rated their own emotional experience. We will not analyze the condition in which participants reported what most people would feel. Second, All data exclusion details specified in the original preregistration remain applicable, with two exceptions: (1) we will exclude reaction times shorter than 200 ms; (2) Because the model estimates separate parameters for specific combinations of response type and normativity, reliable estimation requires a sufficient number of observations in each modeled cell. Therefore, if a participant had too few observations in one or more modeled cells, for example in the cell corresponding to counter-normative unpleasant responses, the model may not converge properly. We will not impose a fixed exclusion threshold in advance. Instead, if convergence problems arise, we will inspect the cell-specific trial counts to determine whether they are attributable to an extremely small number of observations in a particular condition (~2-3). In such cases, participants with insufficient observations in the relevant modeled cells will be excluded from the modeling analysis. We will apply this criterion while aiming to maintain proportional exclusions across countries. Third, we will also analyze the perceptual task. For this analysis, we will apply the same logic of data exclusion used for the emotional task, but we will not exclude additional participants beyond those excluded in the emotional-task analysis. Thus, the perceptual-task analysis will include only participants who were retained in the emotional-task analysis. Fourth, we will also fit a more complex model than the one described in the original preregistration. In this model, the decision threshold and starting point will be defined as a function of the response option (pleasant and unpleasant). Drift rate and drift-rate variability will be defined as a function of stimulus pleasantness (pleasant or unpleasant), and response normativity (normative versus counter-normative). We will set the scale by fixing one of the population-level sv parameters and will use model priors based on the posterior means obtained in the replication model reported by Berkovich and Meiran (2024). The same modeling logic will be applied to the perceptual task. Specifically, the starting point and decision threshold will be defined as a function of the response option (yellow versus blue). Drift rate and drift-rate variability will be defined as a function of the stimuli (blue versus yellow), and accuracy. It should be noted that the stimulus range in the perceptual task differs from the stimulus range in the emotional task. Therefore, if necessary (for example in cases of poor model fit/implausible parameter estimates/ or other indications that the model does not adequately capture the data), we will trim trials with an almost equal ratio of yellow and blue dots. As prios, we will use the most relevant and up-to-date posterior available from previous studies that used tasks as similar as possible to the present perceptual task. If no sufficiently comparable posterior information is available, we will use non-informative priors to avoid biasing parameter estimation. Model fit will be evaluated using either RMSEA or posterior predictive checks. At this stage, we are not yet certain which of these approaches will be most suitable for assessing model fit both at the overall sample level and separately within each country. Therefore, we will first examine the feasibility and suitability of both approaches for this purpose, and will then select one of them as the final method for evaluating model fit. The selected approach will be used to assess whether the model adequately captures the observed data both in the full sample and within each country-specific subsample. In the modeling of the emotional task, it will also be necessary to define response norms. These norms will be defined separately for each country, based on the affective ratings. Finally, we will also analyze the affective-rating task itself. Specifically, we will examine whether the mean and standard deviation of each affective category, pleasant and unpleasant, differ across countries, with each category analyzed separately.These analyses include only the images that were actually entered into the model for each country. The aim of this project is to examine in which components of the affect-labeling process cultural differences emerge. More specifically, there are logical reasons to expect a possible effect of country on each of the model parameters. Therefore, this analysis is exploratory in the sense that we allow for the possibility of country differences in all parameters and their subcomponents. Because these hypotheses are somewhat exploratory in nature, we will adopt a conservative criterion in which only Bayes factors greater than or equal to 10 will be taken as evidence supporting H1.